Wow! I didn't even know a 28 MHz STE accelerator existed!

Thanks! It looks complicated, but it's a pretty straightforward pin-to-pin circuit. Ribbon cable would have been much less messy, but I didn't think of that. XD

Well, there's not a lot of room in the computer, and I was trying to figure out a way to fit it all in without cutting holes in anything. A cable-less circuit would also be a lot harder to design, due to the different ways the corresponding pins of each CPU form are oriented.

I tried looking for such a solution, but everything was either not made anymore or ridiculously expensive. Although my version doesn't work at the moment, I made it using materials I already had, except for the PLCC plug which was really hard to find. I finally found one place that sells 68-pin male PLCC plugs for $30. The accelerator itself cost me $30 (a buy-it-now eBay auction that I was the first to see ), but I'm sure it's worth much more than that, so an additional $30 to try and fit it in my STE was worth it to me. Also, adapters like the ones you linked to are pretty tall, especially if the accelerator were plugged in. I'm not sure if they'd fit, since the STE's CPU is under the keyboard.

Maybe something like this: http://www.adapt-plus.com/products/adap … pplcc.html

Hardcore cable mess. Kudos!
As for the problem, it might have to do with the cables indeed/ Can't you make something that avoids the cabling?

I bought a brand new boxed "ICD Adspeed 16" 16 MHz accelerator on eBay last summer for $30 shipped. It's supposed to replace the CPU in an STF. I could put it in my stock STF, but I'd rather put it in my STE (which I've already upgraded to 4MB RAM and TOS 2.06). STF's and STE's use the same CPU - the only problem is that the STF's CPU comes in a rectangular 64-pin DIP form, while the STE's CPU comes in a square 68-pin PLCC form. So a while ago over winter break I made my own CPU adapter:

Don't laugh! It's my first ever attempt at circuitboard etching, and I didn't have the patience to figure out a circuit design program, so I hand drew the circuits with Sharpie and nail polish. I'm not the neatest at soldering either.

Here it is installed in my STE, with the STF accelerator in the socket:

There's not a lot of room in the computer so I made 2 boards and connected them with wire. It didn't occur to me to use ribbon cable until afterwards. So that's what it looks like... a little sketchy, but the important thing is that it should theoretically work (corresponding pins properly connected, no short circuits). The problem is that in practice it doesn't work for some reason. I've unplugged and replugged it a few times, but every time I turn on the STE I get either a blank white screen or 20 bombs. I used my multimeter's continuity test to make sure each pair of pins is connected properly, and to make sure there are no short circuits between wrong pins. I'm sure each pin is connected to the correct corresponding pin too. I don't know what's wrong. Anyone have any suggestions? Someone told me it might be that the CPU signals are too weak to travel through the long wires. The chronicles of this project are in a thread on Atari-Forum here.

I've declared the project a failure for now, but I thought you guys might like to see it anyway. I did learn a lot from the attempt though. Maybe I'll give it another shot in the summer. If all else fails I can just stick the accelerator in my STF or sell it. Or perhaps trade with someone who's got an STE accelerator?